The Mining Boom in Context

National Economic Review
National Institute of Economic and Industry Research
No. 67   November 2012

The National Economic Review is published four times each year under the auspices of the Institute’s Academic Board.

The Review contains articles on economic and social issues relevant to Australia. While the Institute endeavours to provide reliable forecasts and believes material published in the Review is accurate it will not be liable for any claim by any party acting on such information.

Editor: Kylie Moreland

© National Institute of Economic and Industry Research

This journal is subject to copyright. Apart from such purposes as study, research, criticism or review as provided by the Copyright Act no part may be reproduced without the consent in writing of the relevant Institute.

ISSN 0813-9474

The mining boom in context
Peter Brain, Executive Director, NIEIR

Abstract
The current mining boom is but the latest in a series of bursts of resource-related activity in Australia. The current attitude seems to be that the boom should be passively accepted as well-deserved good fortune. However, recent experience indicates otherwise. While unquestionably resource booms bring prosperity to the mining industry, as a side-effect they can easily generate recession in non-mining industries, possibly even to the extent that the overall benefit of the boom is questionable. The possibilities are outlined in the present paper.

Mining and resource expansion
Over the past 25 years, an economic literature has developed that focuses on the consequences of a sudden episode of resource expansion/exploitation. The episode may arise from the unexpected discovery of a natural resource or from the rapid exploitation of known resources when an unexpected price rise makes resource extraction highly profitable. Either way, total resource investment rises quickly (i.e. over a year or two) to high levels compared to the long-term historic average. When the investment projects are completed, this is followed by a sharp increase in the rate of growth of resource output.

Such expansions occur most often in the mining sector. Technically, agricultural production is also a resource-based activity but, adjusted for the instability in weather-related drivers, the expansion in agricultural investment and production is usually stable – the most recent exception would be the wool boom, which occurred during the Korean War in the early 1950s. Apart from such occasions as the wool boom, sudden and large increases in resource activity are restricted to mineral and energy natural resources (coal, oil and gas), where new discoveries and/or large price changes that improve the economics of past discoveries can trigger sharp increases in investment claims on national economic capacity. In the present discussion, mining expansion is used interchangeably with resource development.

How an episode of resource expansion develops through time
Figure 1 describes the four periods of an episode of elevated resource development. In the period before elevated activity commences levels of investment and output growth are below their long-term historic averages. In this period, new discoveries of mineral resources are made and/or there is a sharp increase in mineral and energy commodity prices, which increase the prospective return on investment. This is followed by the construction or investment stage, where the rate of resource investment as a percentage of GDP increases to well above the long-term average.

The completion of the investment projects ushers in a period that is characterised by high rates of growth in mining production compared to the long-term average growth rate or (at least) the growth rates of the stable periods. During this period, as production expands, real mineral prices generally fall, resulting in falling resource investment. Where a resource expansion occurs based on a single discovery, investment will fall even if mineral prices remain high because of a shortage of unexploited deposits that can be extracted economically. However, in a country like Australia with an extensive catalogue of charted but undeveloped resources, the downturn will not occur until resource prices fall.

In period four, the episode of resource expansion ends with investment and output growth rates returning to below historic benchmarks.

Capture1

The term ‘mining boom’ most commonly refers to the second period, when both investment and real mineral prices are above historical averages. It is only in the fourth period, after all the dust has settled, that it will be possible to fully assess the benefits and costs of the boom.

The Australian experience: Characteristics of episodes of elevated mining expansion
Following on from Figure 1, the characteristics of episodes of mining expansion can be described by the outcomes for investment and mining output growth. Figure 2 shows the level of net mining investment since 1978 as measured by the change in the real mining capital stock in place (see Australian Bureau of Statistics cat. 5204). Since 1978, there have been three episodes of elevated mining expansion. The first episode covered the 1980s, while the second ran from 1995 to 1997. The third episode commenced in 2004 and is currently ongoing. A total of 7 years elapsed between the end of the construction phase for episode one and the commencement of the production phase for episode two. Another 7 years separated the end of the construction phase for episode two and the commencement of the construction phase for episode three.

The current construction episode is likely to continue until at least 2015. Projects under construction, committed and highly likely to proceed will keep the net mining investment average over the next 5 years in the vicinity of A$33 to A$38 billion. In its March 2011 bulletin ‘Australian Commodities’, ABARES predicted that the volume of mining production will grow by 6.2 per cent per annum between 2010 and 2015. This growth rate is common for the energy minerals (coal, LNG and oil) and for iron ore, although not necessarily for all other minerals.

Capture2

The production growth profile is consistent with the immediate past and immediate future level of mining investment. Between 1979 and 2010, an average CVM$1 million of net mining investment produced CVM$0.34 million of mining gross product. (CVM = chain volume measure, which is essentially a means by which the Australian Bureau of Statistics converts values to constant-price terms, in this case prices centred on 2008.) Therefore, an average of CVM$30 billion of investment (or the average from 2007 to 2012) would be expected to produce approximately CVM$10 billion of mining gross product. This represents 7 per cent of the estimated 2011–2012 mining gross product.

The projected growth rate over the next 5 years is not as large as the average annual growth rate from 1984 to 1990, which was 8.5 per cent per annum. This is also reflected in the profile given in Figure 3, which shows the annual average growth rate over the previous 20 quarters (i.e. 5 years). Between the June quarter 1985 and the September quarter 1992, the average annual growth rate exceeded 6 per cent per annum.

Figure 3 also profiles the series for the mining gross product growth rate weighted by the share of mining gross product in GDP or the direct contribution of mining to national GDP growth. Over the production period of the first episode, the average annual contribution to GDP growth was 0.7 percentage points. The contribution over the next 5 years will average 0.6 percentage points, which will be close to the 1980 decade production outcome. Although average annual growth rates will be lower over the next few years than in the 1980s production period, the mining sector now has a larger share of GDP than it had in the mid-1980s.

The quarterly series for national gross mining investment as a share of non-primary product (i.e. excluding agricultural and mining gross product) is shown in Figure 4.

In the early 1980s, the share was approximately 2 per cent. This fell to 1 per cent by the end of the decade, before recovering to the 2-per cent benchmark by 1997 and then falling to 0.8 per cent of GDP by the end of the 1990s. By the middle of the last decade, the 2-per cent benchmark had been regained. Currently, the level of mining investment is running at a little under 4 per cent of GDP and is expected to remain within the 4 to 5-per cent range over the next 5 years.

Capture3 Capture4

From Figure 5, the share of replacement investment in gross mining investment averaged two-thirds of the total gross mining investment over the period 1979 to 2010. Over recent years, however, the share of replacement investment in total gross investment has averaged approximately 40 per cent. It should be noted that the quarterly series for mining investment excludes items that are included in the annual ‘Australian National Accounts’ (cat. no. 5204) series. The quarterly series is approximately 80 per cent of the annual series.

An indicator of the intensity of mining activity (IMA) can be derived by adding the four quarter span growth rate of mining gross product to the mining investment share in non-primary GDP. (The four-quarter span growth rate is a moving average, which for the June quarter approximates the growth rate from one financial year to the next.) This is shown in Figure 4. The figure shows that the intensity indicator currently, and into the immediate future, will take values that are unprecedented compared to outcomes over the past three decades. This partly reflects the fact that mining capital stock has increased three-fold since the late 1980s to reach CVM$304 billion by 2010, with a proportionate increase in replacement investment. Deleting replacement investment means that, to date, the current episode of resource expansion is adding approximately 2.0 to 2.5 percentage points to the IMA. This works out at annual average net investment of $33 billion across period two of the current expansion, which represents 2.8 per cent of current non-primary gross product at factor cost or 2.2 per cent in terms of the quarterly series.

Capture5

The drivers of episodes of mining expansion

The drivers of expansion for resource-based industries are different from drivers in other industries and very different from those in manufacturing industries, where both types of industry are subject to overseas competition. The following discussion does not apply to the small proportion of either manufacturing or mining  industries  that  receive  ‘natural  protection’.

Some of the naturally-protected manufacturing industries, such as gold smelting, are closely tied to mine sites because they greatly reduce the bulk of the product to be transported; others, such as baking, are closely tied to consumption sites because of the costs of transporting fresh products. Similarly, some types of ‘mining’ (chiefly blue-metal quarrying) receive natural protection due to the heavy weight and low value of the product. These cases are discounted, and, henceforth, the designations ‘mining’ and ‘manufacturing industry’ both refer to industries that do not receive natural protection.

As Figure 6 indicates, the drivers of a mining expansion are standard market signals. Typically, an increase in demand forces up mineral prices, which, in turn, not only signals the need for expansion but provides the cash flow to finance expansion. Investment increases are sustained until the supply response drives the price level back to the cost of the next, for example, new mine, LNG plant or transport facility. However, an episode of mining expansion can also occur in response to discoveries without the inducement of an increase in real commodity prices. Hence, the catalyst is the availability of economically extractable resources at prevailing commodity prices.

Capture6

The drivers of manufacturing expansion
Figure 6 also shows the drivers of expansion for manufacturing. Relative costs are important in the sense that manufacturing will contract if there is too great a gap between domestic and foreign costs of production. However, even if relative costs are comparable and Australian products have a price edge (as when the actual $A/$US exchange rate is below its purchasing power parity level), manufacturing expansion depends on producers’ ability to gain a competitive edge by product differentiation in terms of, for example, the design, functionality and durability of their products. This requires years of lead time in:

  1. R&D efforts;
  2. marketing efforts; and
  3. financing innovation and new capacity involving the latest technology.

 

The efforts of a firm in terms of adopting best practice production technology, innovation via R&D expenditures and market development expenditures are all part of either achieving competitive edge product differentiation or identifying opportunities for greater exploitation of existing advantages.

For this type of manufacturing, the individual producer creates or maintains a market while for mining the producer responds to the market. This is why differentiated product manufacturing is riskier than most other industries. An important aspect of this higher level of risk is that differentiated product manufacturers have to create their own finance for expansion, whereas in mining this finance is delivered by the market.

In addition to mining and manufacturing, agriculture (with the partial exceptions of fresh milk and fresh vegetables), tourism and, increasingly, education and health services are counted among the trade-exposed industries. The agricultural industries were akin to mining in that they produced standard commodities with world market prices but they are becoming similar to manufacturing in that they are increasingly developing specialised and individually-marketed products for niche markets. Tourism, education and health produce services rather than goods, but, like manufacturing, serve differentiated markets that must be cultivated assiduously. This study concentrates on manufacturing but its results can be extended to other trade-exposed industries whose product-development requirements (see Figure 6) resemble manufacturing rather than the strict commodity responses characteristic of mining.

At the macroeconomic level, the different drivers of mining versus manufacturing expansion can lead to a conflict between manufacturing expansion and equivalent mining expansion that is unrelated to issues of national resource availability. This is because the higher terms of trade associated with mining expansion are generally followed by an increase in the exchange rate, which makes manufacturing activity less profitable. The converse negative impact on mining from manufacturing expansion is much weaker because manufacturing expansion does not influence the terms of trade.

The most important dynamic is one of cumulative causation. Success in sustained manufacturing expansion depends on an uninterrupted sequence of steps that are resourced adequately and are consistent with market requirements.

Recent past episodes of mining expansion
Figure 7 compares the mining intensity indicator with Australian real non-rural commodity prices in $US, where the price relativity indicator is the Australian consumer price index. The first and third episodes correspond to the dynamics depicted in Figure 6. In the middle of the construction phase for the first episode, the real commodity price indicator was around unity; it fell to 0.7 when the production phase of the first episode ended.

At the end of the second episode, the real price indicator had fallen to 0.75. However, in the early stages of the construction phase of the third episode, the real price indicator had reached values of 1.1 to 1.2. By the end of 2010, the real price indicator value was 1.7. The exception to the rule was the generally low commodity prices that prevailed over the construction phase for the second (relatively subdued) episode. However, the current episode is following the general script. The high current values for the mining intensity indicator compared to the first and second episodes reflect the current relatively high real commodity prices compared to past episodes.

Figure 8 shows the relationship between real commodity prices in $US at the market exchange rate between the Australian dollar and the $US, divided by the purchasing power parity $US exchange rate. In this series, a ratio above unity indicates that the Australian dollar is overvalued compared to the exchange rate required for cost parity between Australia and the United States. We expect that the Australian exchange rate will tend to be overvalued at times of high real commodity prices. As expected, the current high real commodity prices are producing an overvalued exchange rate. At the end of 2010, the overvaluation was 50 per cent and the extent of the overvaluation increased into the second quarter of 2011.

An important point also shown in Figure 8 is that the appreciating currency leaves the mining sector with substantial real price gains in Australian dollar terms. Even when real commodity prices are deflated by the exchange rate over/undervaluation index, real commodity prices are currently higher than those that prevailed during the construction phase of the first episode of mining expansion.

Capture7 Capture8 Capture9

Periods of highly overvalued exchange rates associated with elevated mining activity intensity are very destructive for manufacturing. This is because high relative costs, in conjunction with already high risks, lead producers to curtail or end new development initiatives. R&D is scaled back and capacity expansion and replacement decisions are postponed, which leads to producers falling further behind their competitors in other economies. When the period of elevated mining expansion ends and the exchange rate falls back to cost parity levels, domestic competitors are too far behind to restart R&D programs or even in some cases to undertake the replacement investment required to ensure long-term business sustainability. The same adjustment process occurs, although less severely in terms of the long-run negative outcomes, for other trade-exposed industries such as differentiated agriculture, high-value business services industries, tourist industries and the health and education industries.

Under market conditions, therefore, the dynamics of mining expansion are likely to produce a permanent contraction in manufacturing and other trade-exposed industries compared to what would otherwise have been the case. Each period of elevated resource expansion has a cost in terms of these crowding-out or displacement effects. Each episode of elevated mining expansion produces increased import shares and/or stagnant relative export levels that are not reversed when the period of elevated mining expansion ends.

An additional factor is the pressure on labour resources. The lower the unemployment rate and especially the higher the utilisation of skilled construction labour, the more likely labour will be attracted away from non-resource trade-exposed sectors to mining and related construction. This is particularly likely to be the case in the investment phase. This disrupts the capacity expansion process for non-resource industries, which will not be fully restored when the high mining investment phase ends and labour utilisation rates fall.

Prima facie evidence for this crowding-out or displacement dynamic would be a high negative correlation between the manufacturing share of GDP and the IMA indicator. However, given the dynamics outlined above, the expected negative correlation is not between the manufacturing share in GDP and the IMA, but between the manufacturing share in GDP and the cumulative IMA, or perhaps the cumulative IMA less replacement investment. This is because the greater the intensity of an episode of mining expansion, the greater the permanent reduction in manufacturing capacity and capability. Thus, the time-series outcome for the manufacturing share in GDP should be highly correlated with the cumulative impact of each episode of mining expansion if the above relative industry dynamics have validity.

The data is presented in Figure 9. The strong correlation is self-evident. The correlation coefficient is −0.99. This also implies that the net gains from mining expansion could be small or negative.

An alternative interpretation of Figure 9 is that the tariff phase-down slimmed manufacturing to competitive levels and released resources for mining. To counter this interpretation, the impact of mining expansion on the metals and machinery (MM) manufacturing industry is examined over the past quarter century. The MM sector was not much affected by the tariff phase-down and, more importantly, would be expected to directly benefit from episodes of elevated mining expansion in the form of increased orders during the construction phase of mining expansion.

Metals and machinery manufacturing during recent mining booms
Figures 10 to 13 present a range of MM sector indicators. The figures refer to domestic demand, meaning that exports have been excluded. The evidence from the figures supports our account of manufacturing growth dynamics in that each episode of elevated mining expansion with its accompanying overvalued exchange rate has increased the import share in domestic demand and, importantly, this share is not recovered during subsequent periods of low commodity prices/exchange rates and low IMA values. This means that for each subsequent episode of elevated mining expansion, the domestic MM sector has less capacity available to support the mining expansion with local content. Clearly, the damage done to non-resource trade-exposed sectors of the economy in terms of this crowding out or displacement from episodes of mining expansion is cumulative.

Two terms are used in the literature to describe this process: the ‘Dutch disease’ and the ‘resource curse’.

The steel sector
The quality of Australian manufacturing data has declined over recent years. For example, after the June quarter 2009, quarterly sales data by three-digit Australian and New Zealand Standard Industrial Classification level is no longer available. The data from the MM sector, given above, is based on the Australian National Accounts aggregate data.

The ‘steel sector’ data, or iron and steel plus fabricated metals, is based on the now discontinued data updated to 2010.4 as best as possible. The capacity series estimates shown in the table are based on the traditional trend through peak method plus an 18 per cent loading to bring the series in line with the survey estimates of capacity utilisation.

Capture10 Capture11 Capture12 Capture13

Figures 14 to 16 give the capacity estimates and the capacity utilisation rates. Across the steel industry, capacity utilisation rates have fallen considerably since 2008. Both steel industries currently have an utilisation rate of approximately 50 per cent compared to a normal level of approximately 75–80 per cent.

The profile for the import penetration for the steel industries is shown in Figure 17 and for the machinery sector in Figure 18. If 2003–2004 is taken as the benchmark, the trend increase in import penetration into the steel industry subtracted approximately 6 percentage points from capacity utilisation rates. An approximate 4 to 6 per cent of capacity utilisation loss in the steel sector has come from the increase in import share in the machinery sector. The steel sector lost orders from the machinery sector as import penetration increased in the latter sector.

The increase in import penetration explains approximately half the loss in capacity utilisation rates over the past 3 years. A large part of the rest would probably be explained by the decline in steel-intensive construction, such as offices and apartment buildings.

The core point from the steel sector changes is that the current capacity utilisation rates are low by historical standards, with substantial risk that unless something is done to remedy this situation, a substantial part of current capacity will be permanently closed over the next few years, inflicting significant damage on the economy. The damage will become painful once mining investment and exchange rates start to fall and the unemployment rate and current account deficit start to rise and domestic capacity is no longer available to substitute for imports that can no longer be afforded.

The Dutch disease and the resource curse
The term ‘Dutch disease’ was originally coined by The Economist in 1977 to describe what had happened to the Dutch economy and, in particular, it’s manufacturing sector after the discovery of a large natural gas resource in the late 1950s.

In the early 1980s, economists developed formal models to describe the operational impacts of the Dutch disease, typically a three-sector model comprising:

  1. a resource sector, generally mining;
  2. a non-resource tradable sector (agricultural/ manufacturing/tourism); and
  3. a non-tradable services sector.

The discovery and exploitation of large-scale cost-competitive mineral resources at a time of worldwide supply shortages, as reflected in high real commodity prices, will lead, especially in a small open economy, to:

  1. large-scale capital inflows and rapid growth in mining investment; and
  2. appreciation of the currency and reallocation of resources away from the non-resource sectors and, in particular, the non-resource tradable sector towards the mining and construction sectors.

The competitiveness and capacity of the non-resource tradable sector declines. To some economists this is not a problem. They argue that countries should specialise in industries where they have a comparative/ competitive advantage. The crowding out of non-resource tradable industries is part and parcel of economies maximising their living standards through greater specialisation in what they can (now) do best.

The designation of the Dutch disease, however, describes a case where, in the longer run, productivity and employment would have been higher in the absence of an intense episode of resource development. Clearly, this will be the case if the resource runs out within a decade or two, as was the case of natural gas in the Netherlands. To its credit, the Netherlands Government realised this before it was too late and took action to gain general benefits from its burst of offshore gas production.

When the mineral resource base goes into decline, an expansion of the non-resource tradable sector is required to offset the decline. However, this cannot be easily done because, during the years of resource expansion, declines in non-resource investment, R&D and skill formation widen the competitive gap between the sector and its (previous) foreign peers. Neither cash flow nor institutional support measures are available to help close the gap. As a result, trend growth will decline and per capita GDP levels and living standards (consumption per capita) will fall below the levels that would have been achieved in the absence of the episode of resource development and production.

Capture14 Capture15 Capture16 Capture17 Capture18

For countries with large reserves of unexploited resources that cannot be exhausted in the foreseeable future, the concept of the Dutch disease has been extended to cover the net outcome once the investment/expansion phase has ended and the production phase commenced, bringing with it a supply response. In this case, the crowding-out effect is caused by the high exchange rate and high utilisation of skilled labour during the construction phase of the episode of elevated resource expansion, with the Dutch disease occurring if manufacturing output contracts from what would otherwise have been the case. The need to reinterpret the Dutch disease along these lines for the Australian case is clear. The original Dutch disease event referred to a once-off resource expansion event with the product supply (gas) from the investment coming to an end. In the Australian case, the natural resource base of the economy has allowed sustained long-term expansion. In addition, cost is associated with the episodes of concentrated investment and associated supply growth.

In this context, and adopting the three-sector model framework, Table 1 lists the options that could follow an episode of high resource investment. The table refers to the production impacts of the resource supply after the period of elevated investment and the associated high commodity prices, exchange rates and skilled labour supply pressure has ended. Unless there is a skilled labour supply constraint, overall growth exhibiting positive net additionality is virtually guaranteed during the construction phase. The doubt is whether positive net additionality carries over into the production phase.

Table 2 sets the criteria that apply to resource expansion outcomes. A resource curse outcome applies when there is little or no net addition to overall growth.

The label ‘Dutch disease’ applies (irrespective of whether or not there is overall additional growth) if activity in the manufacturing sector (or, more broadly, the non-resource tradable sector) declines from levels that would otherwise have been achieved in the absence of the episode of elevated mining expansion and the decline is proportional to the expansion in resource production.

Although the definitions in the table refer to the post-investment phase, the outcomes for the non-resource sectors in the economy will largely depend on what happens during the investment/construction phase of the resource expansion. This is because, as outlined above, episodes of elevated resource investment can all too easily result in the non-resource trade-exposed sector being crowded out, expressed in terms of long-term declines in capacity installed compared to what otherwise would have been the case. These impacts have long lags and it may be well into the production phase of an episode of elevated resource investment before the negative production consequences flowing from the investment phase are realised.

It can be seen from Table 2 that there is one case where both the Dutch disease and resource curse apply. The case of a resource curse without the Dutch disease mainly applies to developing economies without a substantial manufacturing sector. In the case of an economy like Australia’s, with a significant manufacturing sector, if the resource curse applies it will most likely be associated with the occurrence of a severe case of the Dutch disease.

Conclusion
An analytical structure for assessment of the benefits of episodes of enhanced mining expansion has been developed in this article. It is undisputed that such episodes generate prosperity for the mining industry; the question is to what extent these benefits are offset by decline in other industries and, if so, over what time periods. The worrying possibility is that a boom in mining investment will divert investment resources away from non-mining investment to the extent that the non-mining industries cannot recover after the mining boom is over. The result will be general prosperity during the boom followed by an accentuated slump. These possibilities await practical investigation in a later article.

References

Australian Bureau of Statistics (2000), Australian National Accounts, cat. no. 5204, Australian Bureau of Statistics, Canberra.

Capture19 Capture20

 

 

 

 

 

 

 

The current mining boom is but the latest in a series of bursts of resource-related activity in Australia. The current attitude seems to be that the boom should be passively accepted as well-deserved good fortune. However, recent experience indicates otherwise. While unquestionably resource booms bring prosperity to the mining industry, as a side-effect they can easily generate recession in non-mining industries, possibly even to the extent that the overall benefit of the boom is questionable. The possibilities are outlined in the present paper.